Research Article  Open Access
G. M. Ji, T. Kanstad, Ø. Bjøntegaard, "Crack Risk Evaluation of Submerged Concrete Tunnel during Hardening Phase", Advances in Civil Engineering, vol. 2018, Article ID 7354025, 14 pages, 2018. https://doi.org/10.1155/2018/7354025
Crack Risk Evaluation of Submerged Concrete Tunnel during Hardening Phase
Abstract
Cracking of concrete structures during the hardening phase often seriously compromises not only structure integrity but also durability and longterm service life. Especially for large massive structures, for example, concrete submerged tunnel, the reliable crack risk evaluation at the hardening phase is critical to the successful design. Mineral additives such as silica fume (SF), blast furnace slag (BFS), and fly ash (FA) have been used extensively in production of highperformance concrete in the last decades. The mineral additives such as FA and BFS not only reduce the hydration heat during the hardening phase but also have significant influence on the development of mechanic and viscoelastic properties at an early age. The main objective of the research is to propose a design methodology to select the appropriate composition of concrete for construction of the submerged tunnel. The influence of mineral additives such as FA and BFS on the risk of cracking during the hardening phase was investigated for the massive concrete structure. Five types of concrete mixes denoted as SV40, 40% BFS, 60% BFS, 40% FA, and 60% FA concrete are considered in the current study, and the measurement to reduce the initial temperature is also considered for 60% FA concrete. First, the welldocumented material models are verified by calibration of restraint stress development in the TSTM test by using the finite element method (FEM), and then the 3D thermalstructural analysis is performed to assess the cracking risk for the submerged tunnel during the hardening phase. Based on analysis results, the 60% FA concrete has both the lowest maximum temperature and the lowest stress/strength ratio, and the crackingfree design based on the current study ensures the successful construction of the submerged tunnel.
1. Introduction
In the past, prediction of the earlyage cracking was almost exclusively based on temperature criteria. The temperature development in the young concrete was calculated, and cracking was predicted from the maximal temperature difference in the massive concrete structure. To avoid cracking, limitations were applied to maximum temperature and temperature difference between the surface and the center of the structure and between the new and the older adjoining structures. These limitations were based on practical experience and experience from the laboratory. The main drawback of the temperaturebased crack risk estimation is that the other important factors in stress calculation are not considered: restraint conditions, material properties, and shrinkage. Several researchers [1, 2] have shown that there is no general correlation between stresses and temperature. Whether young concrete will crack or not depends very much on restraint conditions and material properties.
In North America, the second Midtown Tunnel built under the Elizabeth River from 2013 to 2016 is the first deepwater concrete immersedtube tunnel and only the second allconcrete immersed tunnel in the U.S. The allconcrete tunnel design allows for a strong, durable structure with substantial economic savings compared to a more conventional design using a steel tube encased in concrete, and it is extensively used across Europe. The allconcrete tunnel design is also selected for the submerged tunnel built in Oslo.
Cracking during the hardening phase is prone to occur in massive concrete structures, and it compromises not only structure integrity but also durability and longterm service life. The earlyage cracking was observed in concrete bridge deck expansion joint repair sections [3]. The cracking risk at the hardening phase is the main concern in the design of the concrete submerged tunnel built in Oslo. The illustration of the submerged tunnel is shown in Figure 1. The total length of the tunnel is about 1110 m with three traffic lanes in each direction, and 675 m of it (30 sections) is submerged under the seawater.
(a)
(b)
In recent years, an increased interest in cracking of hardening concrete has led to extensive research on this subject [4–9]. A large number of material models for young concrete have been presented and implemented in computer programs for the simulation of stress development. Simulation of the hardening structure in general has to take into account temperature development due to hydration, development of material properties, and restraint conditions of the particular structure [10, 11]. FEM simulation was performed to predict early thermal stress in second lining concrete of NATM tunnels, and the model was verified by field measurements [12]. The effect of reinforcement on the earlyage cracking is investigated [13], and the results showed that the probability of cracking in highly reinforced structures is lower than that has been estimated in calculations where the effect of reinforcement is completely ignored. For quantifying the effect of reinforcement on the stress development due to restrained loadindependent deformations, a “strain enhancement factor” has been introduced. From the experiments with HSC specimens reinforced with four rebars, this factor reached a mean value of 1.85. A simplified model is used to investigate the possible effect of a gradient concrete material distribution in mass concrete structures on crack reduction. The results of the analysis show that gradient concrete might contribute to lowering the constraint stresses and therefore the crack risk during concrete hardening [14]. The influence of gravel (consisting mainly of quartz), basalt, granite, and limestone aggregate on the temperature development, stress level, and cracking risk has been studied in the experimental and numerical tests [15]. The use of aggregates with appropriate thermal properties in concrete such as low specific heat, high thermal conductivity coefficient, and low coefficient of thermal expansion reduces the induced stresses and the cracking risk. The reinforcement concrete structure was studied by several authors by using software DIANA, ABAQUS, ANSYS, etc. [16–18].
In the current study, a design methodology is proposed based on both comprehensive tests and advanced numerical simulations. The procedure of the design methodology is as follows:(i)Suggest candidate concrete with different compositions(ii)Establish material models through a comprehensive test program(iii)Calibrate material models against temperaturestress testing machine (TSTM) tests(iv)Perform advanced thermalstructural numerical simulation(v)Recommend the concrete composition with lowest cracking risk
2. Candidate Concrete
The concrete proposed in the current design includes one typical construction concrete (SV40) and four other concrete types with different percentages of mineral additives (FA or BFS), and the composition of concrete is presented in Table 1. The materials tests and the mechanical properties do not include the steel reinforcement (bars, etc.).

3. Material Models
A comprehensive test program was performed at laboratories of Norwegian Public Roads Administration (SVV) and Norwegian University of Science and Technology (NTNU) to determine the following material parameters [19]:(i)Heat of hydration(ii)Mechanical properties (elastic modulus, compressive strength, and tensile strength)(iii)Creep/relaxation properties under compressive loading(iv)Volume change
3.1. Heat of Hydration
A semiadiabatic temperature test was performed to determine the hydration heat and thermal properties. The heat of hydration is expressed in the simple threeparameter equation, which is commonly used in engineering practice [20]:where is the asymptotic value of the produced heat, while and are model parameters which are determined from experimental data (Table 2). The activation energy is determined by strength development at several isothermal temperature histories. For temperature calculation, the most important input parameters, in addition to the produced heat and heat capacity, are the heat conductivity and the boundary conditions (connectivity and ambient air temperature) at the various surfaces of the structure.
 
^{1}Binder = cement + silica fume + fly ash + blast furnace slag. 
3.2. Mechanical Properties
The modified version of CEBFIP MC 1990 is used to describe the development of the compressive strength, tensile strength, and modulus of elasticity [21]:in which is introduced to identify the start of significant mechanical properties development, and it might be determined from a TSTM test; it is the time at which stiffness achieves a value high enough to produce measurable stresses. The parameter was determined from the compressive strength development, whereas parameters and were determined from the tensile strength and elastic modulus tests, respectively (Table 3).

3.3. Creep in Compression
The creep of concrete at the constant moisture and thermal state may be well described by power curves of load durations and by inverse power curves of age at loading. This leads to the most wellknown compliance function double power law [22, 23]:
The double power law (DPL) is originally proposed for hardened concrete. It was modified to describe the creep property of earlyage concrete by taking into account the aging characteristics of young concrete [24, 25]:where = compliance function, = concrete age at loading (days), = concrete age (days), = modulus of elasticity at loading time, and , , and = creep model parameters.
From the previous study [19], it can be seen that the combined compressive and tensile creep data give best correspondence with the measured stress developments in the TSTM tests, and using only the compressive DPL parameters gives better agreement with the test results than using only the tensile DPL parameters. For the FEM program DIANA which is used in the current study, only one set of DPL parameters could be specified in the input file, and then the compressive creep data as shown in Table 4 are used in the analysis.

3.4. Volume Change
The thermal dilation (TD) and the autogenous deformation (AD), which are the driving forces to the restraint stress, were determined from the dilation rig test [26]. The sum of the two properties is measured directly and then separated into TD and AD. But this separation is only valid for that particular temperature history. The practical solution is simply to assume that the thermal dilation coefficient is a constant, and then the autogenous deformation is determined by subtracting thermal dilation from the total deformation measured under one realistic temperature history.
The imposed temperatures and measured total deformations for five concrete types are shown in Figure 2. The constant coefficients of thermal expansion (CTEs) used to separate AD and TD and autogenous deformation (AD) after 12 days are presented in Table 5. The separated AD and TD are also shown in Figure 2.
(a)
(b)
(c)
(d)
(e)
(f)

4. Verification of Material Models
Stress development in restraint specimens exposed to realistic temperature histories was measured in the temperaturestress testing machine (TSTM) for all five concrete types. A numerical study was performed by the 3D finite element program DIANA [27] to verify the test results.
The calculated stress developments of concrete containing 40% and 60% BFS have very good agreement with the test results, and the maximum deviation in both compression and tension is about 0.2 MPa (Figure 3). The calculated compressive stress of the concrete containing 40% and 60% FA with 20°C initial temperature is higher than the measured ones, and the calculated tensile stress of the 60% FA concrete is lower than the test results, while the calculated tensile stress of the 40% FA concrete is lower than the test results at first 10 days; afterwards, it becomes higher than the test results. The calculated compressive stress and tensile stress of the 60% FA concrete with 11°C initial temperature are higher than the test results. The calculated compressive stress of the SV40 concrete is higher than the test results, and the calculated tensile stress is lower than the test results until 14 days.
(a)
(b)
(c)
(d)
(e)
(f)
All the model parameters used in the analysis were determined from independent tests, and no parameters are adjusted to achieve better fit with the test results. It can be concluded that the model parameters used in the analysis make reasonably accurate prediction of the stress development for the concrete containing different percentages of FA and BFS, but the deviation for the SV40 concrete is high especially for the tensile stress at 3 days; this is probably due to the low elastic modulus of the SV40 concrete from the test. The maximum compressive and tensile stresses are summarized in Table 6 with the maximum deviation of compressive and tensile stresses. The deviation between the calculated and measured maximum stresses after 12–17 days is less than 12%, and in most cases, the calculations overestimate the maximum tensile stresses.

5. Numerical Simulation
In the design process, the temperature and stress development in the submerged tunnel have been predicted by thermalstructural analysis to assess the risk of through cracking in the hardening phase. The numerical simulation provides reliable insight into the temperature and stress evolution in massive concrete from the casting to 28 days when the properties of concrete are stabilized, and it is essential for the environmental and cost friendly design of the submerged concrete tunnel.
5.1. Numerical Method
The detailed numerical method is described in [28]. In the FE analyses, the thermalstructural problem is decoupled and solved in sequence by the finite element program DIANA. The temperature distribution over time is solved first, and these results are used as input for the subsequent stress calculation. The temperature gradient mainly depends on the total quantity of hydration heat, boundary conditions, thermal properties, and discontinuity in geometry and material properties. The stress gradient depends on temperature distribution, mechanical properties, restraint conditions, discontinuity in geometry and material properties, etc. It is more convenient to simulate the mass structure without reinforcement, and the results are conservative. In the current study, the steel reinforcement (bars, etc.) is not included in the numerical model.
Stress calculations need finer mesh than temperature calculations, the element in stress analysis has to be of higher order than the element in temperature analysis, and the requirements of stress analysis are usually decisive for element selection. The compatibility of the element type used in temperature and stress analysis is automatically handled by DIANA.
5.2. Finite Element Modelling
The tunnel consists of 50 sections, and the length of each section is 22 m. The cross section of the tunnel is not exactly symmetric, and the width between left side and middle wall is slightly different from the width between right side and middle wall. In the current study, the cross section is treated as symmetric, only half of the section is used in numerical analysis, it is considered that the simplification reduces the elements used in the finite element analysis (FEA) by 50%, and analysis results are still accurate enough for the design purpose. The boundary condition at the symmetry plane is fixed in the horizontal direction and has no heat transformation. The main dimensions of the tunnel structure are shown in Figure 4. The element mesh of the 3D model used in the analyses is shown in Figure 5. Due to symmetry conditions, only onefourth of the structure between the dilation joints is modelled. The bottom slab is modelled as hardened concrete, while it is assumed that the walls and top slab are cast in one operation.
The boundary conditions used in analysis are the following:(i)Fresh concrete temperature: 20°C(ii)Ambient air temperature: 20°C(iii)Wind velocity: 1 m/s(iv)Vertical walls (21 mm plywood formwork; convectivity: 0.0033 kJ/m^{2}s°C)(v)Top slab (plastic foil; convectivity: 0.0076 kJ/m^{2}s°C)(vi)Time for formwork and plastic foil removal: 7 days (convectivity: 0.0133 kJ/m^{2}s°C)
The 20node solid element CHX60 is used to model the concrete, and the element is automatically converted to 8node HX8HT in the heat analysis. The 4node boundary element BQ4HT is used to model the boundary conditions in the heat analysis.
6. Analysis Results and Discussion
The typical temperature and stress contour distribution in the middle section of the tunnel is shown in Figures 6 and 7 for SV concrete, and all the other four concrete types have similar temperature and stress contour distribution. The maximum temperature appears at the corner between the inner wall and the top slab, while the critical locations, regarding the risk of through cracking determined as the ratio between maximum tensile stress and tensile strength, are in the center of the inner wall and approximately 0.6–1.2 m above the foundation slab. For the design of the submerged tunnel, the cracking in the outer wall is most critical to structure integrity and functionality, and the discussion of cracking risk will focus on the outer wall. In general, risk needs to take into account the hazard, exposure, and vulnerability over the time period. In the current study, the cracking risk is evaluated quantitatively, and the cracking index is defined as the ratio between stress and tensile strength of concrete. The maximum temperatures in outer and middle walls, in addition to the maximum temperatures in the whole structure, are presented in Table 7.
(a)
(b)
(a)
(b)

The temperature and stress developments for 40% and 60% FA and 40% and 60% BFS concrete are shown in Figures 8–12. The calculated tensile strength after 2 weeks is presented in Table 8, and the corresponding stress/strength ratios at 14 days are summarized in Table 9 for the five types of concrete.
(a)
(b)
(a)
(b)
(a)
(b)
(a)
(b)
(a)
(b)


The results show that the risk of cracking is about 6–14% higher in the inner wall compared to the outer wall due to the higher degree of restraint in the inner wall, but the consequences of through cracking are most serious in the outer wall which therefore should have to meet the design crack criterion. It is seen that, for the initial temperature of 20°C, concrete with 60% FA achieves both the lowest maximum temperature (42.2°C/temperature rise 22.2°C) and the lowest risk of cracking (0.74) in the outer wall. The replacement of cement with FA or BFS reduces the maximum temperature and temperature rise significantly. The temperature rise at the critical location of the outer wall during hardening is 40.7°C for SV40 concrete, 26.5°C for 40% FA concrete, 25.6°C for 40% BFS concrete, and 24.9°C for 60% BFS concrete. The comparison of the crack index at the outer wall for different concrete is shown in Figure 13 and Table 9. Controlling of the initial temperature is the effective method to reduce the cracking risk for the massive concrete structure. The reduction of the fresh concrete temperature from 20 to 11^{o}C results in a reduction of maximum temperature of 5.7°C, while the stress/strength ratio is reduced from 0.74 to 0.52 in the outer wall.
7. Conclusion
In the design of the massive concrete structure, the cracking risk in the hardening phase is the most critical for the integrity and durability of the structure. The proposed design methodology ensures that the appropriate composition of concrete is selected for the construction of the concrete submerged tunnel. The numerical method proposed here is based on the finite element method, and it has the potential to be used as a design tool in the important submerged concrete tunnel project.
The risk of through cracking for the concrete submerged tunnel is determined for five types of concrete mixes denoted as SV40, 40% FA, 60% FA, 40% BFS, and 60% BFS concrete. The maximum temperature appears at the corner between the inner wall and the top slab, while the critical locations, regarding the risk of through cracking determined as the ratio between maximum tensile stress and tensile strength, are at the center of the inner wall and approximately 0.6–1.2 m above the foundation slab. The temperature development is similar in the critical locations of the inner and outer walls, but the stresses and the risk of cracking are 12% higher in the inner walls due to a larger degree of restraint. However, only the outer wall will experience water pressure, and therefore, the results of the outer wall are given most attention. It is seen that concrete with 60% FA has both the lowest maximum temperature (42.2°C/temperature rise 22.2°C) and the lowest risk of cracking (0.74) in the outer wall. The temperature rise during hardening is 30.7°C for SV40 concrete, 26.5°C for 40% FA concrete, 25.6°C for 40% BFS concrete, and 24.9°C for 60% BFS concrete. Furthermore, it is seen that the reduction in fresh concrete temperature from 20°C to 11°C results in a reduction of maximum temperature of 5.7°C, while the stress/strength ratio is reduced from 0.74 to 0.52.
Data Availability
The data used in the paper are accessible through the website https://brage.bibsys.no/xmlui/handle/11250/236401.
Conflicts of Interest
The authors declare that there are no conflicts of interest regarding the publication of this paper.
Acknowledgments
The financial contribution of the Norwegian Research Council is gratefully acknowledged. The NORCRACK partners were the Norwegian University of Science and Technology (project leader), Skanska Norge ASA, Elkem ASA Materials, Norcem AS, Fesil ASA, and the Norwegian Public Roads Administration. The paper is mainly based on the PhD project performed by Guomin Ji [19].
References
 S. Bernander and M. Emborg, “Risk of Cracking in Massiv Concrete StructuresNew Development and Experiences,” in Proceedings of the RILEM International Symposium on Thermal Cracking in Concrete at Early Ages, R. Springenschmid, Ed., pp. 385–392, E & FN Spon, London, UK, 1994. View at: Google Scholar
 M. Emborg, “Thermal stresses in concrete structures at early ages,” Luleå University of Technology, Luleå, Sweden, 1989, Doctoral thesis. View at: Google Scholar
 J. R. Wright, F. Rajabipour, J. A. Laman, and A. Radlińska, “Causes of early age cracking on concrete bridge deck expansion joint repair sections,” Advances in Civil Engineering, vol. 2014, Article ID 103421, 10 pages, 2014. View at: Publisher Site  Google Scholar
 P. Havlásek, V. Šmilauer, K. Hájková, and L. Baquerizo, “Thermomechanical simulations of earlyage concrete cracking with durability predictions,” IOP Conference Series: Materials Science and Engineering, vol. 236, 2017. View at: Publisher Site  Google Scholar
 A. B. E. Klausen, “Early age crack assessment of concrete structures: experimental determination of decisive parameters,” Ph.D. thesis, NTNU, Trondheim, Norway, 2016, ISBN 9788232618507. View at: Google Scholar
 IPACS, “BriteEuram project BRPRCT970437,” Improved Production of Advanced Concrete Structures, 19972001. View at: Google Scholar
 J. Olofsson and M. Uhlán, Round Ronbin Simulation–Ground Slab Examples, Skanska Teknik AB, Stockholm, Sweden, 2000, ISBN 9189580540.
 R. Springenschmidt, Thermal Cracking in Concrete at Early Ages, Proceedings of the International RILEM Symposium, E&FN Spon, London, UK, 1995.
 R. Springenschmidt, R. Breitenbucher, and M. Mangold, Development of the Cracking Frame and Temperature Stress Testing Machine, Thermal Cracking in Concrete at EarlyAge, EF Spon, London, UK, 1994.
 A. B. E. Klausen, T. Kanstad, Ø. Bjøntegaard, and E. Sellevold, “Comparison of tensile and compressive creep of fly ash concretes in the hardening phase,” Cement and Concrete Research, vol. 95, pp. 188–194, 2017. View at: Publisher Site  Google Scholar
 A. B. E. Klausen, T. Kanstad, and Ø. Bjøntegaard, “Updated TemperatureStress Testing Machine (TSTM): introductory tests, calculations, verification and investigation of variable fly ash content,” in Proceedings of International Conference: Concreep10 Mechanisms and Physics of Creep, Shrinkage, and Durability of Concrete and Concrete Structures, p. 9, Vienna, Austria, September 2015. View at: Google Scholar
 C. K. Rankoth, A. Hosoda, and K. Iwama, “Modeling and verification of early age thermal stress in second lining concrete of NATM tunnels,” Journal of Advance Concrete Technology, vol. 15, no. 6, pp. 213–226, 2017. View at: Publisher Site  Google Scholar
 M. S. Sule and K. van Breugel, “Effect of reinforcement on earlyage cracking in high strength concrete,” HERON, vol. 49, no. 3, 2004. View at: Google Scholar
 E. Strieder, R. Hilber, E. Stierschneider, and K. Bergmeister, “FEstudy on the effect of gradient concrete on early constraint and crack risk,” Applied Sciences, vol. 8, no. 2, p. 246, 2018. View at: Publisher Site  Google Scholar
 B. Klemczaka, M. Batogb, M. Pilcha, and A. Zmij, “Analysis of cracking risk in early age mass concrete with different aggregate types,” in Proceedings of International Conference on Analytical Models and New Concepts in Concrete and Masonry Structures AMCM’2017, 2017. View at: Google Scholar
 J. Fu, Y. Wu, and Y.b. Yang, “Effect of reinforcement strength on seismic behavior of concrete moment frames,” Earthquake and Structures, vol. 9, no. 4, pp. 699–718, 2015. View at: Publisher Site  Google Scholar
 B. Hu, J. Wang, and G. Li, “Numerical simulation and strength models of FRPwrapped reinforced concrete columns under eccentric loading,” Construction and Building Materials, vol. 25, no. 5, pp. 2751–2763, 2011. View at: Publisher Site  Google Scholar
 A. I. Karabinis, T. C. Rousakis, and G. E. Manolitsi, “3D finiteelement analysis of substandard RC columns strengthened by fibre reinforced polymer sheets,” Journal of Composites for Construction, vol. 12, no. 5, pp. 531–540, 2008. View at: Publisher Site  Google Scholar
 G. M. Ji, “Cracking risk of concrete structures in the hardening phase: experiments, material modelling and finite element analysis,” NTNU, Dept. of Structural Eng., Norway, 2008, Doctoral thesis. View at: Google Scholar
 P. F. Hansen and E. J. Pedersen, “Maturity computer for controlled curing and hardening of concrete,” Nordisk Betong, vol. 1, no. 19, pp. 21–25, 1977. View at: Google Scholar
 T. Kanstad, T. A. Hammer, Ø. Bjøntegård, and E. J. Sellevold, “Mechanical properties of young concrete: Evaluation of test methods for tensile strength and modulus of elasticity,” 1999, Determination of model parameters, SINTEFreport no. STF22 A99762, Trondheim. View at: Google Scholar
 Z. P. Bažant and L. Panula, Simplified Prediction of Concrete Creep and Shrinkage From Strength and Mix, Structural Engineering Report, No. 7810/6403, Department of Civil Engineering, Technological Institute, Northwestern University, Evanston, IL, USA, 1978.
 Z. P. Bažant and L. Panula, “Practical prediction of time dependent deformations of concrete,” Materials and Structures, Part I and II, vol. 11, no. 5, pp. 307–328, 1978. View at: Publisher Site  Google Scholar
 T. Kanstad, T. A. Hammer, Ø. Bjøntegaard, and E. J. Sellevold, “Mechanical properties of young concrete: Part 1:Experimental results related to test method and temperature effects,” Materials and Structures, vol. 36, no. 258, pp. 218–225, 2003. View at: Publisher Site  Google Scholar
 T. Kanstad, T. A. Hammer, Ø. Bjøntegaard, and E. J. Sellevold, “Mechanical properties of young concrete: Part 2: Determination of model parameters and test program proposals,” Materials and Structures, vol. 36, no. 4, pp. 226–230, 2003. View at: Publisher Site  Google Scholar
 Ø. Bjøntegaard, T. A. Hammer, and E. J. Sellevold, “On the measurement of free deformation of early age cement paste and concrete,” Cement and Concrete Compositions, vol. 26, no. 5, pp. 427–435, 2004. View at: Publisher Site  Google Scholar
 DIANA FEA User's manual, 2017, DIANA FEA BV.
 G. M. Ji, T. Kanstad, and Ø. Bjøntegaard, “Numerical modelling of field test for crack risk assessment of early age concrete containing fly ash,” Advances in Materials Science and Engineering, vol. 2018, Article ID 1058170, 16 pages, 2018. View at: Publisher Site  Google Scholar
Copyright
Copyright © 2018 G. M. Ji et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.